Updated naming convention for dataset databases to be clearer.

Fixed bugs in the dataload process with referencing the new json folder and then looking up the available databases.

Updated the dataloader to ignore case on product features. Note have deliberately not harmonised product feature capitalisation or similar features as this "messiness" is representative of real organisational data.

Fixed a bug in generate data where it didn't reference the json folder in the new file sructure

Added the scripts which were used to build the dataset to the repo, and tweaked to use common code